Due to their manual, error-prone, and resourceintensive nature, traditional network management methods become unfeasible for highly dynamic, complex, and heterogeneous Beyond 5G (B5G) systems. In this landscape, Intent-Based Networking (IBN) emerges as a promising solution that automates the service lifecycle while offering flexibility and scalability. Nevertheless, IBN faces challenges in expressing, interpreting, and refining high-level intents across diverse Use Cases (UCs). This work proposes a Language Model Intent-based Application Programming Interface (LM-IbAPI) for enhanced dynamic network management in B5G infrastructures to address these challenges. In the framework of the 6GDIFERENTE project, our solution is seamlessly integrated with the management layer to translate multiple intents from diverse UCs into precise network actions with accuracy and reduced time response. We employed Ollama as a unified framework to quickly and effectively test several language models. The results demonstrated that mistral_7B achieved the best trade-off between processing time and accuracy, being capable of generalising across varying numbers of users, services, operators, and South-bound Interfaces (SBIs).

In recent years, wireless networks are evolving complex, which upsurges the use of zero-touch artificial intelligence (AI)-driven network automation within the telecommunication industry. In particular, network slicing, the most promising technology beyond 5G, would embrace AI models to manage the complex communication network. Besides, it is also essential to build the trustworthiness of the AI black boxes in actual deployment when AI makes complex resource management and anomaly detection. Inspired by closed-loop automation and Explainable Artificial intelligence (XAI), we design an Explainable Federated deep learning (FDL) model to predict per-slice RAN dropped traffic probability while jointly considering the sensitivity and explainability-aware metrics as constraints in such non- IID setup. In precise, we quantitatively validate the faithfulness of the explanations via the so-called attribution-based log-odds metric that is included as a constraint in the run-time FL optimization task. Simulation results confirm its superiority over an unconstrained integrated-gradient (IG) post-hoc FDL baseline.

One of the current challenges still facing artificial intelligence (AI) and machine learning (ML) based solutions is their performance in real life environments. In this work, we report and analyze the outcomes of the deployment of one of our previously proposed AI-based automatic tracking areas (TA) design algorithm in a real and live mobile network. The implemented automation is based on our solution in [1], and has been experimented in an area with accidental land morphology that often leads to aggressive overshooting, and other radio issues affecting many high traffic sites. The area under consideration includes one major city, surrounding suburbs and villages, and main road traffic sections. We first analyze the characteristics of the considered area of deployment, collect the required statistics in terms of handover attempts, mobility event measurement reports (MRs), total paging requests per source site, and the intersite distances ISD. We then build a real dataset and construct the needed similarity matrix to run our ML-based algorithm. Finally, we present and compare the results collected after the deployment of our proposed solution in the live network with the initially existing TA plans. The obtained reduction in overhead costs of both tracking area update and paging signaling exceeded the expectations and the MNO targets

The evolution of Software-Defined Networking (SDN) and Network Function Virtualization (NFV) in the telecommunications industry have intensified the issues of network management at large scales. Dynamic service orchestration and adaptive resource allocation became a necessity for network operators to manage the rapid growth of users and data-intensive applications. The impact of network automation on energy consumption and overall operating costs is often overlooked. Guaranteeing strict performance constraints of Ultra-Reliable Low Latency Communication (URLLC) services while enhancing energy efficiency is one of the major critical problems of future communication networks, given the urgency to reduce carbon emissions and energy consumption. In this work, we study the problem of zero-touch Service Function Chain (SFC) orchestration for multi-domain networks, targeting the latency reduction of URLLC services while improving energy efficiency for beyond-5G networks. Specifically, we propose SCHE2MA, a Service CHain Energy-Efficient Management framework based on distributed Reinforcement Learning (RL), that can intelligently deploy SFCs with shared VNFs per se into a multi-domain network. Finally, we evaluate SCHE2MA through model validation and simulation while demonstrating its ability to jointly reduce average service latency by 103.4% and energy consumption by 17.1% compared to a centralized RL solution.

eration (6G)-enabled massive network slicing is a strong enabler for the expected pervasive digitalization of the vertical market. In such a context, artificial intelligence (AI)-driven zero-touch network automation should present a high degree of scalability and sustainability, especially when deployed in live production networks wherein the collected monitoring datasets at different points are non-independent and identically distributed (non-IID). This paper presents a new cloud-native service-level agreement (SLA)-driven stochastic policy to guarantee a scalable and fast operation of constrained federated learning (FL)-based analytic engines that perform statistical slice- level resource provisioning at RAN-Edge in a non-IID setup. Both simulated and cloud-native emulated scenarios are implemented to demonstrate the superiority of the solution in reducing SLA violation, convergence time and computation cost compared to different FL baselines, showcasing thereby a higher scalability.

A key enabler for both scalability and sustainability in beyond 5G (B5G) network slicing consists on minimizing the exchange of raw monitoring data across different domains. This is achieved by bringing the analysis functions closer to the data collection points. To this end, we introduce in this paperstatistical federated learning(SFL) provisioning models that can learn over a live network non independent identically distributed (non-IID) datasets in an offline fashion while respecting slice-level service level agreement (SLA) long-term statistical constraints. Specifically, we consider three resource SLA metrics, namely,cumulative distribution function(CDF),$Q$-th percentileandmaximum/minimum bounds. These metrics are dataset-dependent and non-convex non-differentiable and, to sidestep the inaccuracy of settling only for surrogates, we propose a novel formulation that jointly considers the statistical objective and constraints as well as their smooth approximation using theproxy-Lagrangianframework, which we solve via a non-zero sum two-player game strategy. Numerical results on various slice-level resources show that SFL enables SLA enforcement while significantly reducing the overhead compared to both state-of-the-art FedAvg and centralized constrained deep learning schemes. Finally, we provide an analysis for the lower bound of the so-calledreliable convergence probabilityin the SFL setup.

Scalability and sustainability are the corner stones to unleash the potential of beyond fifth-generation (B5G) ultra-dense networks that are expected to handle massive and heterogeneous services. This implies that the transport of the underlying raw monitoring data should be minimized across the network, and urges to bring the analysis functions closer to the data collection points. While federated learning (FL) is an efficient tool to implement such a decentralized strategy, real networks are generally characterized by time- and space-varying users distributions, traffic profiles, and channel conditions. This makes the data collected across different points non independent and identically distributed (non-IID), which is challenging for FL tasks. To cope with this issue, we first introduce a new a priori metric that we call dataset entropy, whose role is to capture the distribution, the quantity of information, the unbalanced structure and the “non-IIDness” of a dataset independently of the models. This entropy is calculated using a clustering scheme based on a similarity matrix defined over both the features and the supervised output spaces, and is targeting classification as well as regression tasks. The FL aggregation server then uses the reported dataset entropies to devise i) an entropy-based federated averaging scheme, and ii) a stochastic participant selection policy to significantly stabilize the training, minimize the convergence time, and reduce the corresponding computation cost. Numerical results are provided to illustrate all these advantages.

Artificial intelligence (AI)-driven zero-touch massive network slicing is envisioned to be a disruptive technology in beyond 5G (B5G)/6G, where tenancy would be extended to the final consumer in the form of advanced digital use-cases. In this paper, we propose a novel model-free deep reinforcement learning (DRL) framework, called collaborative statistical Actor-Critic (CS-AC) that enables a scalable and farsighted slice performance management in a 6G-like RAN scenario that is built upon mobile edge computing (MEC) and massive multiple-input multiple-output (mMIMO). In this intent, the proposed CS-AC targets the optimization of the latency cost under a long-term statistical service-level agreement (SLA). In particular, we consider the Q-th delay percentile SLA metric and enforce some slice-specific preset constraints on it. Moreover, to implement distributed learners, we propose a developed variant of soft Actor-Critic (SAC) with less hyperparameter sensitivity. Finally, we present numerical results to showcase the gain of the adopted approach on our built OpenAI-based network slicing environment and verify the performance in terms of latency, SLA Q-th percentile, and time efficiency. To the best of our knowledge, this is the first work that studies the feasibility of an AI-driven approach for massive network slicing under statistical SLA.

In this paper, we address the concept of dynamic resource allocation for radio access network (RAN) slicing in beyond 5G (B5G) systems under service-level agreement (SLA). Using live network distributed key performance indicators (KPIs) mini-datasets, we introduce a new class of federated learning models that can capture the long-term cumulative distribution function (CDF) statistic—is usually used to define SLA—and enforce some preset constraints on it. Given that the CDF is also dataset-dependent and non-convex non-differentiable, we formulate the corresponding local optimization task using the proxy-Lagrangian framework and solve it via a non-zero sum two-player game strategy. Numerical results show that the proposed decentralized resource allocation approach enables SLA enforcement and significantly reduces the SLA violation rate for various slice-level KPIs.

To harness the full potential of beyond 5G (B5G) communication systems, zero-touch network slicing (NS) is viewed as a promising fully-automated management and orchestration (MANO) system. This paper proposes a novel knowledge plane (KP)-based MANO framework that accommodates and exploits recent NS technologies and is termed KB5G. Specifically, we deliberate on algorithmic innovation and artificial intelligence (AI) in KB5G. We invoke a continuous model-free deep reinforcement learning (DRL) method to minimize energy consumption and virtual network function (VNF) instantiation cost. We present a novel Actor-Critic-based NS approach to stabilize learning called, twin-delayed double-Q soft Actor-Critic (TDSAC) method. The TDSAC enables central unit (CU) to learn continuously to accumulate the knowledge learned in the past to minimize future NS costs. Finally, we present numerical results to showcase the gain of the adopted approach and verify the performance in terms of energy consumption, CPU utilization, and time efficiency.

Artificial intelligence (AI)-driven zero-touch network slicing (NS) is a new paradigm enabling the automation of resource management and orchestration (MANO) in multi-tenant beyond 5G (B5G) networks. In this paper, we tackle the problem of cloud-RAN (C-RAN) joint slice admission control and resource allocation by first formulating it as a Markov decision process (MDP). We then invoke an advanced continuous deep reinforcement learning (DRL) method called twin delayed deep deterministic policy gradient (TD3) to solve it. In this intent, we introduce a multi-objective approach to make the central unit (CU) learn how to re-conFigure computing resources autonomously while minimizing latency, energy consumption and virtual network function (VNF) instantiation cost for each slice. Moreover, we build a complete 5G C-RAN network slicing environment using OpenAI Gym toolkit where, thanks to its standardized interface, it can be easily tested with different DRL schemes. Finally, we present extensive experimental results to showcase the gain of TD3 as well as the adopted multi-objective strategy in terms of achieved slice admission success rate, latency, energy saving and CPU utilization.

In this paper, we investigate the concept of OPEX-limited resource provisioning as a key component in fifth generation (5G) radio access networks (RAN) slicing. The different RAN slices' tenants (i.e. logical operators) are dynamically allocated isolated portions of physical resource blocks (PRBs), baseband processing resources and backhaul capacity. To achieve this dynamic resource allocation, we rely on key performance indicators (KPIs) datasets stemming from a live cellular network endowed with traffic probes. These datasets are used to train a new class of deep neural networks (DNNs) models where OPEX requirements, formulated as non-convex non-differentiable violation rate constraints, are also dataset-dependent. The designed constrained DNNs are then optimized via a non-zero sum two-player game strategy. In this respect, we highlight the effect of the different hyperparameters on the respect of the OPEX limitations, while ensuring a dynamic RAN resource orchestration that follows the slices' traffics trends.

In this paper, we address the issue of resource provisioning as an enabler for end-to-end dynamic slicing in software defined networking/network function virtualization (SDN/NFV)-based fifth generation (5G) networks. The different slices’ tenants (i.e. logical operators) are dynamically allocated isolated portions of physical resource blocks (PRBs), baseband processing resources, backhaul capacity as well as data forwarding elements (DFE) and SDN controller connections. By invoking massive key performance indicators (KPIs) datasets stemming from a live cellular network endowed with traffic probes, we first introduce a low-complexity slices’ traffics predictor based on a soft gated recurrent unit (GRU). We then build—at each virtual network function—joint multi-slice deep neural networks (DNNs) and train them to estimate the required resources based on the traffic per slice, while not violating two service level agreement (SLA), namely,violation rate-based SLA andresource bounds-based SLA. This is achieved by integrating dataset-dependent generalized non-convex constraints into the DNN offline optimization tasks that are solved via a non-zero sum two-player game strategy. In this respect, we highlight the role of the underlying hyperparameters in the trade-off between overprovisioning and slices’ isolation. Finally, using reliability theory, we provide a closed-form analysis for the lower bound of the so-calledreliable convergence probabilityand showcase the effect of the violation rate on it.

Reliability and latency challenges in future mixed sub-6 GHz/millimeter wave (mmWave) fifth generation (5G) cell-free massive multiple-input multiple-output (MIMO) networks is to guarantee a fast radio resource management in both uplink (UL) and downlink (DL), while tackling the corresponding propagation imbalance that may arise in blockage situations. In this context, we introduce a semi-blind UL/DL decoupling concept where, after its initial activation, the central processing unit (CPU) gathers measurements of the Rician K-factor-reflecting the line-of-sight (LOS) condition of the user equipment (UE)-as well as the DL reference signal receive power (RSRP) for both 2.6 GHz and 28 GHz frequency bands, and then train a non-linear support vector machine (SVM) algorithm. The CPU finally stops the measurements of mmWave definitely, and apply the trained SVM algorithm on the 2.6 GHz data to blindly predict the target frequencies and access points (APs) that can be independently used for the UL and DL. The accuracy score of the proposed classifier reaches 95% for few training samples.

In this paper, we address the issue of automatic tracking areas (TAs) planning in fifth generation (5G) ultra-dense networks (UDNs). By invoking handover (HO) attempts and measurement reports (MRs) statistics of a 4G live network, we first introduce a new kernel function mapping HO attempts, MRs and inter-site distances (ISDs) into the so-called similarity weight. The corresponding matrix is then fed to a self-tuning spectral clustering (STSC) algorithm to automatically define the TAs number and borders. After evaluating its performance in terms of the Q-metric as well as the silhouette score for various kernel parameters, we show that the clustering scheme yields a significant reduction of tracking area updates and average paging requests per TA; optimizing thereby network resources.

In this paper, we investigate the performance of multihop communication schemes over millimeter wave Weibull-fading channels. We adopt the so-called detect-and-forward relaying strategy, with generalized M-quadrature amplitude modulations, as it offers an interesting performance-complexity tradeoff. We first provide closed-form and asymptotic expressions for the end-to-end bit error rate (BER) and energy efficiency (EE). Then, we use the obtained results to investigate both BER- and EE-optimal power allocation strategies in the adopted context. The obtained results are assessed through simulations, and the proposed analysis framework provides good insight into the design and optimization of multihop schemes as illustrated with various numerical examples.

In this paper, we address the issue of throughput-efficient half-duplex constrained relaying schemes for broadband uplink transmissions over multiple-input multiple-output (MIMO) channels. We introduce a low complexity signal-level cooperative spatial multiplexing (CM) architecture that allows for the shortening of the relaying phase without resorting to any symbol detection or re-mapping at the relay side. Half-duplex latency is thereby reduced, resulting in a significant throughput gain compared to amplify-and-forward (AF) relaying scheme. Surprisingly, we show that CM strategy becomes more powerful in boosting uplink throughput as the source approaches cell edge.

In this contribution, we present a new class of project-and-forward (PF) relays operating with broadband MIMO hybrid ARQ-aided cooperative systems. Their architecture enables to jointly perform the orthogonal projection over multiple ARQ transmissions, increasing thereby the relay diversity order and achieving an interesting power gain over amplify-and-forward (AF)-based relaying schemes.